mirror of
https://gitlab.com/libeigen/eigen.git
synced 2026-04-10 11:34:33 +08:00
Improved std::complex sqrt and rsqrt.
Replaces `std::sqrt` with `complex_sqrt` for all platforms (previously `complex_sqrt` was only used for CUDA and MSVC), and implements custom `complex_rsqrt`. Also introduces `numext::rsqrt` to simplify implementation, and modified `numext::hypot` to adhere to IEEE IEC 6059 for special cases. The `complex_sqrt` and `complex_rsqrt` implementations were found to be significantly faster than `std::sqrt<std::complex<T>>` and `1/numext::sqrt<std::complex<T>>`. Benchmark file attached. ``` GCC 10, Intel Xeon, x86_64: --------------------------------------------------------------------------- Benchmark Time CPU Iterations --------------------------------------------------------------------------- BM_Sqrt<std::complex<float>> 9.21 ns 9.21 ns 73225448 BM_StdSqrt<std::complex<float>> 17.1 ns 17.1 ns 40966545 BM_Sqrt<std::complex<double>> 8.53 ns 8.53 ns 81111062 BM_StdSqrt<std::complex<double>> 21.5 ns 21.5 ns 32757248 BM_Rsqrt<std::complex<float>> 10.3 ns 10.3 ns 68047474 BM_DivSqrt<std::complex<float>> 16.3 ns 16.3 ns 42770127 BM_Rsqrt<std::complex<double>> 11.3 ns 11.3 ns 61322028 BM_DivSqrt<std::complex<double>> 16.5 ns 16.5 ns 42200711 Clang 11, Intel Xeon, x86_64: --------------------------------------------------------------------------- Benchmark Time CPU Iterations --------------------------------------------------------------------------- BM_Sqrt<std::complex<float>> 7.46 ns 7.45 ns 90742042 BM_StdSqrt<std::complex<float>> 16.6 ns 16.6 ns 42369878 BM_Sqrt<std::complex<double>> 8.49 ns 8.49 ns 81629030 BM_StdSqrt<std::complex<double>> 21.8 ns 21.7 ns 31809588 BM_Rsqrt<std::complex<float>> 8.39 ns 8.39 ns 82933666 BM_DivSqrt<std::complex<float>> 14.4 ns 14.4 ns 48638676 BM_Rsqrt<std::complex<double>> 9.83 ns 9.82 ns 70068956 BM_DivSqrt<std::complex<double>> 15.7 ns 15.7 ns 44487798 Clang 9, Pixel 2, aarch64: --------------------------------------------------------------------------- Benchmark Time CPU Iterations --------------------------------------------------------------------------- BM_Sqrt<std::complex<float>> 24.2 ns 24.1 ns 28616031 BM_StdSqrt<std::complex<float>> 104 ns 103 ns 6826926 BM_Sqrt<std::complex<double>> 31.8 ns 31.8 ns 22157591 BM_StdSqrt<std::complex<double>> 128 ns 128 ns 5437375 BM_Rsqrt<std::complex<float>> 31.9 ns 31.8 ns 22384383 BM_DivSqrt<std::complex<float>> 99.2 ns 98.9 ns 7250438 BM_Rsqrt<std::complex<double>> 46.0 ns 45.8 ns 15338689 BM_DivSqrt<std::complex<double>> 119 ns 119 ns 5898944 ```
This commit is contained in:
@@ -94,19 +94,6 @@ template<typename T> struct scalar_quotient_op<const std::complex<T>, const std:
|
||||
|
||||
template<typename T> struct scalar_quotient_op<std::complex<T>, std::complex<T> > : scalar_quotient_op<const std::complex<T>, const std::complex<T> > {};
|
||||
|
||||
// Complex sqrt is already specialized on Windows.
|
||||
#if EIGEN_COMP_MSVC == 0
|
||||
template<typename T>
|
||||
struct sqrt_impl<std::complex<T> >
|
||||
{
|
||||
EIGEN_DEVICE_FUNC
|
||||
static EIGEN_ALWAYS_INLINE std::complex<T> run(const std::complex<T>& x)
|
||||
{
|
||||
return complex_sqrt<T>(x);
|
||||
}
|
||||
};
|
||||
#endif
|
||||
|
||||
} // namespace internal
|
||||
} // namespace Eigen
|
||||
|
||||
|
||||
@@ -150,7 +150,7 @@ Packet4f prsqrt<Packet4f>(const Packet4f& _x) {
|
||||
|
||||
template<> EIGEN_DEFINE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS EIGEN_UNUSED
|
||||
Packet4f prsqrt<Packet4f>(const Packet4f& x) {
|
||||
// Unfortunately we can't use the much faster mm_rqsrt_ps since it only provides an approximation.
|
||||
// Unfortunately we can't use the much faster mm_rsqrt_ps since it only provides an approximation.
|
||||
return _mm_div_ps(pset1<Packet4f>(1.0f), _mm_sqrt_ps(x));
|
||||
}
|
||||
|
||||
@@ -158,7 +158,6 @@ Packet4f prsqrt<Packet4f>(const Packet4f& x) {
|
||||
|
||||
template<> EIGEN_DEFINE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS EIGEN_UNUSED
|
||||
Packet2d prsqrt<Packet2d>(const Packet2d& x) {
|
||||
// Unfortunately we can't use the much faster mm_rqsrt_pd since it only provides an approximation.
|
||||
return _mm_div_pd(pset1<Packet2d>(1.0), _mm_sqrt_pd(x));
|
||||
}
|
||||
|
||||
|
||||
Reference in New Issue
Block a user